An Evaluation of Messages to Promote Parental Intent to Vaccinate Children Aged <12 Years Against COVID-19

Abstract

Objectives:

Effective health communication can increase intent to vaccinate. We compared 8 messages that may influence parents’ intent to vaccinate their children against COVID-19.

Methods:

In a cross-sectional survey of adults in the United States administered online in August 2021, 1837 parents and legal guardians were exposed to 8 messages (individual choice, gain/practical benefits, nonexpert, health care provider recommendation, altruism/community good, safety/effectiveness, safety, and effectiveness) to determine message reception and influence on intent to vaccinate their children. Parents responded to 10 questions using a Likert scale. We computed odds ratios (ORs) for each message, with an OR >1.0 indicating greater observed odds of participant agreement with the follow-up statement as compared with a reference message. We compared outcomes individually across messages with ordinal logistic regression fit using generalized estimating equations.

Results:

The individual choice message had the highest odds of agreement for understanding intent (OR = 2.10; 95% CI, 1.94-2.27), followed by the health care provider recommendation message (OR = 1.58; 95% CI, 1.46-1.71). The individual choice message had the highest odds of memorability, relatability, and trustworthiness. The altruism/community good message was at or near second best. The altruism/community good message had the highest or near-highest odds of increasing parents’ intent to vaccinate their children, asking friends and family for their thoughts, and searching for additional information. The message that most motivated parents to vaccinate their children depended on parental intent to vaccinate prior to being exposed to the tested messages.

Conclusions:

Messages with themes of individual choice, health care provider recommendation, and altruism/community good may be used in future message campaigns. Further research is needed to refine message concepts related to altruism/community good.

Keywords

vaccines child health health promotion health communications

The Pfizer-BioNTech COVID-19 mRNA vaccine received emergency use authorization (EUA) by the US Food and Drug Administration for children aged 5-11 years on October 29, 2021,¹ and for children aged 6 months to 4 years on June 17, 2022.² Increasing vaccination uptake among all age groups is a critical component of a comprehensive approach to stop the COVID-19 pandemic. Health communication messages are essential to reach key audiences such as parents of young children, a demographic group that has expressed high levels of COVID-19 vaccine hesitancy.³ Multiple aspects of message design, including framing and credibility, are critical to supporting public health vaccination efforts.

Health communication research demonstrates that how a message is framed affects its effectiveness in promoting health behavior change and that the same health information presented in various frames can have different effects on health behavior changes.⁴ A 2012 meta-analysis on message framing found no significant difference in the persuasiveness of gain- and loss-framed appeals for encouraging vaccination.⁵ Results indicated that further testing is necessary for specific audiences because parents might be more persuaded to vaccinate their children by loss-framed appeals than gain-framed appeals.

Additionally, the influence of the individual as compared with the community is integral in message framing. Research has shown how these differences affect health behaviors^6,7 and how people respond to messages.⁸ The underlying feature in these messages is that social norms guide behaviors. Health campaigns that appeal to the individual in places such as the United States, which has a high individualism score,^9,10 succeeded at focusing on changing individual behavior with memorable phrases such as “Back to Sleep” for sudden infant death syndrome prevention¹¹ and “Click It or Ticket” for motor vehicle injury prevention.¹² Health campaigns that appeal to a collectivist worldview emphasize the benefit of changing individual behavior for the health of the community. In the context of COVID-19, collectivism is a potential protective factor: geographic areas with high levels of collectivism had fewer COVID-19 cases and deaths¹³ and higher rates of face mask use than individualistic locations.¹⁴ According to recent COVID-19 vaccination research, vaccination intent was not affected by individual versus collective frames,¹⁵ and priming collectivism—by using the collective “we” rather than the individual “I” in messaging—had increased behavioral intentions toward physical distancing and face mask wearing.¹⁶ Thus, the real impact of individual versus collective message framing is mixed in the context of COVID-19. More research is needed to understand whether these frames may influence parents’ intent to vaccinate their young children against COVID-19.

Testimonials from credible sources or trusted messengers who relay health messages may have strong influences on behavior.^17-19 In the context of vaccination messages, the messenger is a critical aspect of message design and can range from laypeople to perceived experts.²⁰ For COVID-19, messenger expertise is complicated because experts and nonexperts may be trusted sources of information.²¹ Research has established that parents highly trust their health care providers as experts, while trust from other sources (eg, other parents, family members) has also become heavily considered in decision-making.^3,20,22,23 Although health care providers are trusted messengers, parent peers may be equally trusted. More research is needed to better predict which messenger will have the greatest effect on parents’ intent to vaccinate their children against COVID-19.²⁰

Additionally, for message credibility in the context of COVID-19 vaccination, vaccine safety and effectiveness are critical. Because COVID-19 vaccines first operated under EUA and, at the time of this study, were not yet authorized under EUA for children aged <16 years, which includes this study’s sample, the media and parents had questions about vaccine safety and effectiveness.²⁴ In a recent study on message framing for COVID-19 and willingness to get vaccinated, adult participants who received COVID-19 vaccine safety and efficacy messages were more likely than a control group to report intent to get vaccinated.²⁵

We reviewed relevant literature, current data, and ongoing conversations and consultations within the Centers for Disease Control and Prevention’s (CDC’s) COVID-19 Emergency Response Unit to create COVID-19 vaccine health communication messages tailored toward parents who are hesitant to get their children vaccinated. Each message represents specific themes (practical benefits or gains, individual benefits, altruism or community benefits) and credibility factors (expert or nonexpert messenger, safety and/or effectiveness). The objective of this study was to identify messages that positively influence the intent of parents and legal guardians to vaccinate children aged <12 years against COVID-19.

Methods

Researchers administered an online survey to parents and legal guardians in the United States during August 2-13, 2021, before EUA of the COVID-19 vaccine for children aged 5-11 years. Researchers used Qualtrics XM to recruit an online convenience sample through a range of recruitment methodologies to produce a diverse sampling frame; participants received a small incentive. Eligibility criteria included (1) being the parent or legal guardian of ≥1 child or adolescent aged <18 years, (2) having ≥1 child or adolescent aged <18 years who is unvaccinated against COVID-19, and (3) being responsible for making this child’s health care decisions. To aid in rapid data collection and turnaround, the survey asked for no additional geographic or demographic information, except that quotas were established so that approximately half of participants would be parents or legal guardians (hereinafter, “parents”) whose oldest unvaccinated child or adolescent was aged 12-17 years and half whose oldest unvaccinated child was aged <12 years. All participants lived in the United States and were surveyed in English. CDC reviewed this study, and it was conducted consistent with applicable federal law and CDC policy. CDC authorized this research under a waiver from the Paperwork Reduction Act resulting from the COVID-19 public health emergency. All participants provided written consent.

Survey instructions directed participants to consider their oldest unvaccinated child when answering survey questions. At the time of the survey, children and adolescents aged 12-17 years were eligible for COVID-19 vaccination under an EUA by the US Food and Drug Administration,²⁶ while children aged <12 years were not. We present findings on parents whose oldest unvaccinated child was aged <12 years at the time of survey participation and, thus, not yet eligible for vaccination.

In the survey, we tested 8 health communication messages to determine how effective they were at influencing parents’ intent to vaccinate their child (Table 1). Messages appeared randomly to minimize order effect bias.²⁷ Each message included a specific theme or credibility factor. Some messages included scientific data, while others presented first-person narratives.

Table 1.

Message text, follow-up questions, and theme in a survey of messages that can promote parental COVID-19 vaccine intent, United States, August 2-13, 2021

No.	Message text	Message-specific follow-up statement^a	Theme
1	I have the freedom to choose what is best for my children’s health and am choosing to prevent diseases like COVID-19 by getting them vaccinated.	This message makes me feel like I can make my own choice about getting my child vaccinated.	Individual choice
2	Getting my child vaccinated against COVID-19 will help give me more peace of mind for in-person learning, social activities, travel, and sports.	This message makes me feel I will gain something from my child becoming vaccinated.	Gain/practical benefits
3	I will get my children vaccinated against COVID-19 as soon as I can. It’s the best shot they have at getting their childhood back and moving past the pandemic.	This message seems to be coming from another parent like me.	Nonexpert
4	The primary reason my children are getting vaccinated against COVID-19 is because it is recommended by their health care provider, whom I trust.	This message is about following advice from a trusted medical provider.	Health care provider recommendation
5	I’m not only protecting my child but also those around them when I get them vaccinated against COVID-19.	This message makes me feel others are protected when children are vaccinated.	Altruism/community good
6	COVID-19 vaccines may cause my child to have some mild side effects that last a couple of days, but I know they are normal signs my child’s body is building protection against COVID-19.	This message gives me confidence that a COVID-19 vaccine is safe and effective for children.	Safety/effectiveness
7	COVID-19 vaccines have been carefully studied in children, and these studies all indicate the vaccines are safe for them to get.	This message gives me confidence that the COVID-19 vaccine is safe for children.	Safety
8	Study results for children aged 12-15 years showed the vaccine was 100% effective at preventing COVID-19.	This message gives me confidence that the COVID-19 vaccine is effective at preventing COVID-19 and reducing related hospitalization for children.	Effectiveness

Response options were on a scale from 1 to 7 (1 = completely agree, 7 = completely disagree).

After the presentation of each message, the survey asked 3 sets of additional questions. First, the survey asked a message-specific follow-up question, framed as a statement and designed to determine understanding of message intent. For example, the follow-up statement for message 1 (“I have the freedom to choose what is best for my children’s health and am choosing to prevent diseases like COVID-19 by getting them vaccinated”) was “This message makes me feel like I can make my own choice about getting my child vaccinated.” Response options appeared on a Likert-type scale (1 = completely agree, 7 = completely disagree). The next 2 sets of questions appeared after each message. The first repeated follow-up question probed message reception with items for understandability, memorability, relatability, and trustworthiness. It asked, “How much do you agree with the following statements?” The 4 statements were the following: (1) “This message is easy to understand.” (2) “This message is memorable.” (3) “This message was written for parents/caregivers like me.” (4) “I trust the information provided in this message.” This message reception–related set of questions also appeared on a Likert-type scale (1 = strongly agree, 5 = strongly disagree). The second set of repeated follow-up questions, framed as statements, probed influences on vaccination information seeking and decision-making. The question began with “After seeing this message, I will” and provided the following 5 statements: (1) “Have my child vaccinated against COVID-19.” (2) “Talk to my child’s doctor more about if a COVID-19 vaccine is right for my child.” (3) “Ask friends what they think about getting children vaccinated against COVID-19.” (4) “Ask family what they think about getting children vaccinated against COVID-19.” (5) “Search for more information to help me make a decision about getting a COVID-19 vaccine for my child.” Answers to this set of survey items also appeared on a Likert-type scale (1 = definitely will, 5 = definitely will not).

Statistical Analysis

Each participant received all 8 messages. Thus, the 10 Likert scale outcomes were not independent. We compared each of the 10 survey item outcomes individually across messages with ordinal logistic regression fit using generalized estimating equations to account for the dependence.²⁸ The assumed correlation structure was category exchangeable, and we used robust (sandwich) variance estimators.

For each outcome, we computed odds ratios (ORs), 95% CIs, and P values. Each P value tested the null hypothesis that participants would equally endorse a specific follow-up statement across all 8 messages. An OR >1.0 for a given message indicates greater observed odds of participant agreement with the follow-up statement as compared with the reference message. We used message 7 (“COVID-19 vaccines have been carefully studied in children, and these studies all indicate the vaccines are safe for them to get”) as the reference message because it had the lowest odds of being endorsed by respondents. The choice of reference message was arbitrary and did not alter the model fit or conclusions. We tested the significance of ORs by comparing each pair of messages and adjusting for multiple testing across the 28 possible pairs using a Bonferroni correction. Additionally, we conducted 2 exploratory analyses. The first assessed the moderating effects of race and ethnicity, education, urbanicity (ie, urban, suburban, rural), and initial attitude toward vaccinating one’s child to determine if message comparisons differed among levels of these variables. The second assessed whether the message comparisons differed between parents of children aged 0-4 years and 5-11 years. All tests were 2-sided, and we determined significance at the α = .05 level.

An equal distribution of all other variables across the messages occurred because all participants received all 8 messages. Thus, we included no adjustment for confounders in the regression model. We conducted data analysis by using R version 4.0.5 (R Foundation for Statistical Computing) and the multgee package.²⁹

Results

Participant Characteristics

Overall, 1837 participants whose oldest unvaccinated child was aged <12 years completed the online survey (Table 2). Most participants (n = 1738, 94.6%) were parents and the remainder (n = 99, 5.4%) were other legal guardians. Most respondents were aged 25-54 years (n = 1683, 91.6%). About two-thirds were female (n = 1251, 68.1%) and 1223 (66.6%) were married. Most were White (n = 1481, 80.6%) and non-Hispanic/Latino (n = 1628, 88.6%). Fewer than half (n = 875, 47.6%) had a 4-year degree or higher education level, and most were employed full-time (n = 1046, 56.9%). Most lived in an urban (n = 722, 39.3%) or suburban (n = 772, 42.0%) area.

Table 2.

Demographic characteristics of parents of children aged <12 years who had not received the COVID-19 vaccine (N = 1837), in a survey of messages that can influence parental vaccine intent, United States, August 2-13, 2021

Characteristic	Parents of unvaccinated children aged <12 y, no. (%)
Primary health care decision maker
Joint	472 (25.7)
Primary	1365 (74.3)
Age range, y
18-24	113 (6.2)
25-34	799 (43.5)
35-54	884 (48.1)
55-74	41 (2.2)
Sex
Male	582 (31.7)
Female	1251 (68.1)
Prefer not to answer	4 (0.2)
Race (mark all that apply)^a
American Indian/Alaska Native	29 (1.6)
Asian	94 (5.1)
Black or African American	211 (11.5)
Native Hawaiian/Pacific Islander	7 (0.4)
White	1481 (80.6)
Other	60 (3.3)
Prefer not to answer	11 (0.6)
Ethnicity
Hispanic or Latino	209 (11.4)
Not Hispanic or Latino	1628 (88.6)
Sexual orientation
Gay (lesbian or gay)	37 (2.0)
Straight (not gay or lesbian)	1589 (86.5)
Bisexual	168 (9.1)
Something else	32 (1.7)
Don’t know/not sure	11 (0.6)
Marital status
Married	1223 (66.6)
Unmarried, living with a partner	220 (12.0)
Divorced	89 (4.8)
Widowed	18 (1.0)
Separated	43 (2.3)
Single, never been married	241 (13.1)
Prefer not to answer	3 (0.2)
Education
Grade school	10 (0.5)
<High school/some high school	33 (1.8)
High school graduate or GED	358 (19.5)
Some college or technical school	561 (30.5)
Four-year degree	468 (25.5)
Some postgraduate studies	151 (8.2)
Received advanced degree	256 (13.9)
Occupational status
Employed full-time	1046 (56.9)
Employed part-time	216 (11.8)
Unemployed	125 (6.8)
Homemaker	365 (19.9)
Student	18 (1.0)
Retired	21 (1.1)
Disabled	37 (2.0)
Other	9 (0.5)
Type of living environment
Urban	722 (39.3)
Suburban	772 (42.0)
Rural	343 (18.7)

Abbreviation: GED, General Educational Development.

Numbers total to >1837 because respondents could choose >1 response.

Message Comparisons

We found significantly different odds of agreement among follow-up statements related to understanding of message intent (P < .001) (Table 3). Overall, the message on individual choice (message 1) had significantly higher odds of agreement among participants (OR = 2.10; 95% CI, 1.94-2.27; P < .001) when compared with all other messages, after adjusting for multiple comparisons. The health care provider recommendation (message 4), which emphasized credibility, had the second-highest odds of understanding message intent (OR = 1.58; 95% CI, 1.46-1.71; P < .001); it was significantly different from all but the altruism/community good message (message 5) after adjusting for multiple comparisons.

Table 3.

Comparison of messages via ordinal logistic regression, by message follow-up item, in a survey on messages that can promote vaccine intent among parents of children aged <12 years who had not received the COVID-19 vaccine (N = 1837), United States, August 2-13, 2021

Message	OR (95% CI)	P value^a	Pairwise comparison^b
Understanding of message intent ^c (P < .001 ^d )
1. Individual choice	2.10 (1.94-2.27)	<.001	2, 3, 4, 5, 6, 7, 8
2. Gain/practical benefits	1.25 (1.19-1.32)	<.001	1, 4, 5, 6, 7
3. Nonexpert	1.19 (1.13-1.26)	<.001	1, 4, 5, 7
4. Health care provider recommendation	1.58 (1.46-1.71)	<.001	1, 2, 3, 6, 7, 8
5. Altruism/community good	1.57 (1.48-1.67)	<.001	1, 2, 3, 6, 7, 8
6. Safety/effectiveness	1.10 (1.05-1.16)	<.001	1, 2, 4, 5, 7, 8
7. Safety	1 [Reference]	—	1, 2, 3, 4, 5, 6, 8
8. Effectiveness	1.29 (1.23-1.36)	<.001	1, 4, 5, 6, 7
Message reception ^e
Understandability (P < .001^d)
1. Individual choice	1.27 (1.17-1.39)	<.001	3, 6, 7, 8
2. Gain/practical benefits	1.20 (1.11-1.30)	<.001	7
3. Nonexpert	1.07 (0.99-1.17)	.09	1
4. Health care provider recommendation	1.11 (1.02-1.21)	.01
5. Altruism/community good	1.22 (1.12-1.33)	<.001	7
6. Safety/effectiveness	1.07 (0.99-1.16)	.08	1
7. Safety	1 [Reference]	—	1, 2, 5
8. Effectiveness	1.08 (1.00-1.18)	.06	1
Memorability (P < .001^d)
1. Individual choice	1.21 (1.12-1.31)	<.001	2, 4, 6, 7
2. Gain/practical benefits	1.06 (0.99-1.13)	.12	1, 5, 8
3. Nonexpert	1.08 (1.01-1.16)	.03	8
4. Health care provider recommendation	1.03 (0.96-1.10)	.45	1, 5, 8
5. Altruism/community good	1.18 (1.10-1.27)	<.001	2, 4, 6, 7
6. Safety/effectiveness	1.04 (0.98-1.12)	.21	1, 5, 8
7. Safety	1 [Reference]	—	1, 5, 8
8. Effectiveness	1.21 (1.13-1.30)	<.001	2, 3, 4, 6, 7
Relatability (P < .001^d)
1. Individual choice	1.26 (1.16-1.36)	<.001	2, 3, 4, 6, 7, 8
2. Gain/practical benefits	1.10 (1.03-1.18)	.004	1
3. Nonexpert	0.99 (0.93-1.07)	.87	1, 5
4. Health care provider recommendation	1.04 (0.97-1.12)	.23	1
5. Altruism/community good	1.13 (1.06-1.21)	<.001	3, 7
6. Safety/effectiveness	1.02 (0.96-1.09)	.53	1
7. Safety	1 [Reference]	—	15
8. Effectiveness	1.05 (0.98-1.12)	.19	1
Trustworthiness (P < .001^d)
1. Individual choice	1.47 (1.38-1.57)	<.001	2, 3, 4, 6, 7, 8
2. Gain/practical benefits	1.22 (1.15-1.29)	<.001	1, 3, 7, 8
3. Nonexpert	1.07 (1.01-1.14)	.01	1, 2, 4, 5, 6
4. Health care provider recommendation	1.28 (1.20-1.36)	<.001	1, 3, 7, 8
5. Altruism/community good	1.33 (1.25-1.41)	<.001	3, 6, 7, 8
6. Safety/effectiveness	1.19 (1.13-1.27)	<.001	1, 3, 5, 7, 8
7. Safety	1.00 [Reference]	—	1, 2, 4, 5, 6
8. Effectiveness	1.07 (1.01-1.13)	.01	1, 2, 4, 5, 6
Influence on vaccination information seeking and decision-making ^e
Intent to vaccinate child (P < .001^d)
1. Individual choice	1.02 (0.98-1.06)	.26	—
2. Gain/practical benefits	1.05 (1.01-1.09)	.02	6
3. Nonexpert	1.02 (0.98-1.06)	.43	—
4. Health care provider recommendation	1.03 (0.99-1.07)	.16	—
5. Altruism/community good	1.05 (1.01-1.10)	.02	6
6. Safety/effectiveness	0.97 (0.93-1.01)	.11	258
7. Safety	1 [Reference]	—	—
8. Effectiveness	1.05 (1.01-1.09)	.03	6
Talk to child’s doctor about vaccine (P = .35^d)
1. Individual choice	1.03 (0.99-1.08)	.16	—
2. Gain/practical benefits	1.00 (0.95-1.05)	.91	—
3. Nonexpert	1.02 (0.98-1.07)	.35	—
4. Health care provider recommendation	0.99 (0.95-1.04)	.74	—
5. Altruism/community good	1.00 (0.96-1.05)	.94	—
6. Safety/effectiveness	0.99 (0.94-1.04)	.63	—
7. Safety	1 [Reference]	—	—
8. Effectiveness	0.98 (0.94-1.03)	.41	—
Will ask friends for their thoughts (P = .005^d)
1. Individual choice	1.05 (1.00-1.10)	.03	—
2. Gain/practical benefits	1.04 (0.99-1.08)	.11	—
3. Nonexpert	1.02 (0.98-1.07)	.36	—
4. Health care provider recommendation	1.00 (0.95-1.04)	.86	—
5. Altruism/community good	1.06 (1.02-1.11)	.005	6
6. Safety/effectiveness	0.99 (0.94-1.03)	.58	5
7. Safety	1 [Reference]	—	—
8. Effectiveness	1.01 (0.97-1.06)	.60	—
Will ask family members for their thoughts (P = .01^d)
1. Individual choice	1.03 (0.99-1.08)	.14	—
2. Gain/practical benefits	1.04 (0.99-1.09)	.09	4
3. Nonexpert	1.01 (0.97-1.06)	.55	—
4. Health care provider recommendation	0.96 (0.92-1.01)	.10	2, 5
5. Altruism/community good	1.03 (0.99-1.08)	.13	4
6. Safety/effectiveness	0.99 (0.95-1.04)	.70	—
7. Safety	1 [Reference]	—	—
8. Effectiveness	1.02 (0.98-1.07)	.30	—
Will search for additional information (P = .001^d)
1. Individual choice	1.07 (1.02-1.13)	.004	3
2. Gain/practical benefits	1.02 (0.97-1.07)	.40	—
3. Nonexpert	0.98 (0.94-1.03)	.43	1, 5
4. Health care provider recommendation	1.02 (0.97-1.07)	.42	—
5. Altruism/community good	1.06 (1.01-1.11)	.01	3
6. Safety/effectiveness	1.01 (0.96-1.07)	.58	—
7. Safety	1 [Reference]	—	—
8. Effectiveness	1.01 (0.96-1.06)	.70	—

Abbreviation: OR, odds ratio.

Determined by ordinal logistic regression; P < .05 considered significant.

Messages that are significantly different from the message in the row after adjusting for multiple comparisons.

After the presentation of each message, the survey asked a message-specific follow-up question; these questions were framed as statements and designed to determine understanding of message intent. These questions were answered on a Likert-type scale of 1 to 5.

P value for the global 7-df test of the null hypothesis that the 8 messages were endorsed equally.

After the presentation of each message, the survey asked 2 questions in addition to the message-specific follow-up question. The first repeated follow-up question probed message reception (understandability, memorability, relatability, and trustworthiness). The second repeated follow-up question, framed as a statement, probed influences on vaccination information seeking and decision-making. Both sets of questions were answered on a Likert-type scale from 1 to 5.

The message on individual choice also had the highest odds of agreement among participants for the message reception items related to understandability (OR = 1.27), memorability (OR = 1.21), relatability (OR = 1.26), and trustworthiness (OR = 1.47; P < .001 for all) (Table 3). Even after adjusting for multiple comparisons, this message had significantly higher odds of agreement than many of the other messages for these items.

The message on altruism/community good (message 5) consistently scored at or near second best for agreement with the items related to understanding of message intent, understandability, memorability, relatability, and trustworthiness (Table 3). With respect to trustworthiness, the health care provider recommendation (message 4) also performed well (OR = 1.28) but not as well as individual choice (OR = 1.47) and altruism/community good (OR = 1.33; P < .001 for all). Additionally, the health care provider recommendation had significantly higher odds of agreement than the nonexpert message with respect to trustworthiness.

Regarding the 5-item question examining message influence on vaccination information seeking and decision-making, messages had significantly different agreement for 4 of the 5 items. These differences were small in magnitude and generally not significant. For intent to vaccinate, we found significant differences among messages in the group of parents with children aged 5-11 years but not aged 0-4 years (Table 4). While the message concerning individual choice ranked high among all these items, the altruism/community good message had the highest or near-highest odds of increasing intent to vaccinate among the 4 items that had significant message differences: intent to vaccinate child (OR = 1.05; 95% CI, 1.01-1.10; P = .018), will ask friends for their thoughts (OR = 1.06; 95% CI, 1.02-1.11; P = .005), and will search for additional information (OR = 1.06; 95% CI, 1.01-1.11; P = .01); the item “will ask family members for their thoughts” was not significant (OR = 1.03; 95% CI, 0.99-1.08; P = .13). However, for each item, after adjusting for multiple comparisons, responses to this message were not significantly different from almost all the other responses. Additionally, we found no significant differences among messages for the item examining influence on likelihood of talking to the child’s doctor (P = .35).

Table 4.

Results of exploratory analysis of differences in how parents of children aged <5 and 5-11 years perceived messages, in a survey on messages that can promote vaccine intent among parents of children aged <12 years who had not received the COVID-19 vaccine (N = 1837), United States, August 2-13, 2021

Message	Age <5 y (n = 1139, 62.0%)		Age 5-11 y (n = 698, 38.0%)
Message	OR (95% CI)	P value^a	OR (95% CI)	P value^a	Pairwise comparison^b
Influence on vaccination information seeking and decision-making
Intent to vaccinate child	—	.22^c	—	<.001^c	—
Age <5 vs 5-11 y (P for interaction = .02)
1. Individual choice	1.02 (0.96-1.08)	.48	1.02 (0.97-1.08)	.41	—
2. Gain/practical benefits	1.01 (0.94-1.07)	.87	1.08 (1.02-1.14)	.004	6
3. Nonexpert	0.98 (0.92-1.04)	.57	1.04 (0.98-1.10)	.16	—
4. Health care provider recommendation	1.04 (0.97-1.10)	.27	1.02 (0.97-1.08)	.36	—
5. Altruism/community good	1.02 (0.96-1.09)	.49	1.07 (1.01-1.13)	.01	6
6. Safety/effectiveness	0.96 (0.91-1.02)	.22	0.97 (0.92-1.02)	.27	2, 5, 8
7. Safety	1 [Reference]	—	1 [Reference]	—	—
8. Effectiveness	1.01 (0.95-1.07)	.79	1.07 (1.02-1.13)	.01	6

Abbreviation: OR, odds ratio.

Determined by ordinal logistic regression; P < .05 considered significant.

Messages that are significantly different from the message in the row after adjusting for multiple comparisons (only age 5-11 years had any significantly different pairs).

P value for the global 7-df test of the null hypothesis that the 8 messages were endorsed equally.

Our first exploratory analysis found that for race, ethnicity, and education, differences in message comparisons among subgroups were not of large magnitude and did not alter overall conclusions about which messages were best. Initial attitude toward vaccinating one’s child, however, did moderate the message comparison. Which message most influenced intent to vaccinate one’s child depended on initial intent to vaccinate, with greater distinctions between messages among those who had not yet made up their minds or were leaning toward not vaccinating but still open to the idea. Overall, the distinctions between messages with respect to motivating intent to vaccinate were not large. After stratification by initial intent to vaccinate, however, the distinctions were larger (eTable in Supplemental Material). Among those who initially indicated that they will definitely, probably, or definitely not vaccinate their children, the messages that most influenced parents toward greater intent to vaccinate were altruism/community good, gain/practical benefits, and individual choice, respectively, with 19% to 24% greater odds than the least effective message within each stratum. Among those who were unsure or probably would not vaccinate their children, the distinctions were even greater: gain/practical benefits and altruism/community good were the best messages, with 44% and 87% greater odds than the least effective message within each stratum.

In general, our second exploratory analysis found no meaningful differences in how parents of children aged 0-4 years and 5-11 years perceived messages, with 1 exception (Table 4). Parents’ reported intent to vaccinate their children had a significant interaction effect (P = .02), indicating differences in message effect among parents of children aged 5-11 years, despite a much smaller sample size (0-4 years, P = .22, n = 1139; 5-11 years, P < .001, n = 698). Among parents of children aged 5-11 years, the messages concerning gain/practical benefits (OR = 1.08; 95% CI, 1.02-1.14; P = .004), altruism/community good (OR = 1.07; 95% CI, 1.01-1.13; P = .01), and effectiveness (OR = 1.07; 95% CI, 1.07-1.13; P = .01) ranked as best.

Discussion

Messages that focused on individual choice and altruism/community good had the highest odds of participant agreement for positive reactions, comprehension of the intent, and reception (eg, understandability, memorability, reliability, trustworthiness). Of these 2 messages, the individual choice message appealed slightly more, and the altruism/community good message was slightly more effective in increasing vaccination intent. The altruism/community good message may have been perceived as motivating among parents of children aged <12 years because of the lack of EUA for this age group at the time that this study was conducted; the message may have appealed to parents’ sense of being left out of vaccination efforts and the protection provided.^30-32 The message that displayed messenger credibility, the health care provider recommendation, also performed well with respect to participant understanding of message intent and trustworthiness.

Motta et al³³ found influential value in message frames emphasizing collective health consequences on intent to vaccinate. Our study suggests that this effect may also apply to parents’ intent to vaccinate their children. However, Motta et al³³ and others^25,34 found safety, effectiveness, and health risk messaging to be motivating. In contrast, messages to address parents’ concerns about vaccine safety and effectiveness and their children’s risk of illness did not resonate with parents or motivate them in our study. This finding may indicate different motivators when parents consider vaccination of their children versus vaccination of self. Parents may not have been very concerned about safety and effectiveness because children aged <12 years were not yet eligible to be vaccinated. Another possibility is that previous research^25,33,34 did not include a personal choice message, which has become relevant because of differing opinions on the COVID-19 vaccine and intent to vaccinate. Research shows that the source of messaging matters for health communications, and health care provider recommendation is a strong predictor of vaccine uptake in children.^35-38 These survey data show that, in addition to a direct recommendation from one’s own health care provider, a provider recommendation may be effectively used through messages indicating health care endorsement of a vaccine by leveraging credibility and trustworthiness. Health care providers continue to play a key role in supporting COVID-19 vaccination among children aged <12 years and providing vaccination education to parents. Considering the findings of this study and the role of individual choice, it may be especially effective if, while providing a COVID-19 vaccination recommendation, health care providers also emphasize parents’ right to choose vaccination for their children.

Limitations

This survey had several limitations. First, it used a convenience-based sample, which may have introduced selection bias. As such, the sample is not representative of or generalizable to the entire population. However, because each respondent was prompted with all 8 messages, thereby eliminating confounding bias, the message comparisons do have strong internal validity. Generalizability was further limited by the timing of this study. While conducted prior to EUA of the vaccine for use in children aged <12 years, many people knew that the COVID-19 vaccine would be approved for this age group in the near future. Existing research also indicated that safety and efficacy were among many considerations for reaching key audiences of adults for COVID-19 vaccination and parents of children for new vaccines. Yet, conducting this research was valuable to allow public health officials to plan for vaccine promotion upon approval of the vaccine for this age group and to offer insight for future vaccine messaging efforts. Second, measurement issues for the message-specific follow-up items, which were designed to measure understanding of intent (eg, “This message makes me feel like . . .”), limited our study because these were different among the messages and some were double-barreled. These measurement issues may have limited our findings to reactions to a message (vs message understanding) and/or unclear reasons why parents may disagree with a message. Furthermore, because this was a cross-sectional survey, we did not measure subsequent vaccination behaviors. Future messaging research may benefit from clear measurement and longitudinal panels to identify parent-reported feelings, intentions, and vaccination behaviors.

Conclusions

Some parents are hesitant about vaccinating their children against COVID-19, but effective messaging and communications can increase confidence and vaccination rates. Vaccine communication messages should use effective framing and be perceived as credible by parents so that the messages resonate with parents and increase their intent to vaccinate their children. Our study showed that messages emphasizing the fact that vaccinating children helps protect others and the community overall may be the most influential in motivating parental intent to vaccinate. Messages related to the parent’s independent choice in the matter and messages conveying that health care providers recommend the vaccine for children may also be effective. Further research may help enhance and refine precise message concepts that may be effective within the altruism/community good message, such as protecting those who are vulnerable to the severe outcomes of COVID-19 (eg, elderly people).

Supplemental Material

sj-docx-1-phr-10.1177_00333549231218725 – Supplemental material for An Evaluation of Messages to Promote Parental Intent to Vaccinate Children Aged <12 Years Against COVID-19

Supplemental material, sj-docx-1-phr-10.1177_00333549231218725 for An Evaluation of Messages to Promote Parental Intent to Vaccinate Children Aged <12 Years Against COVID-19 by Isabella L. Chan, Kelsey Schwarz, Nicole Weinstein, Gordon Mansergh, Ramzi W. Nahhas, Deborah Gelaude, Robert Alexander, Leslie Rodriguez, Warren Strauss, Torey Repetski, Nancy Sullivan, Everett Long, Steve L. Evener, Adrienne Garbarino and Laura M. Mercer Kollar in Public Health Reports

Footnotes

Acknowledgements

The authors acknowledge all families with children and thank parents and caregivers who are making difficult decisions about their children’s medical care and whether to vaccinate their children against COVID-19. The authors thank the Centers for Disease Control and Prevention Vaccine Task Force Communications leaders, especially Jo Stryker, PhD, and Dayle Kern, MA, who were instrumental in providing support in guiding this research.

Disclaimer

The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this study was provided by the Centers for Disease Control and Prevention’s Vaccine Task Force through contract 75D30121C10149.

ORCID iDs

Isabella L. Chan, PhD

Nicole Weinstein, MSW

Supplemental Material

Supplemental material for this article is available online. The authors have provided these supplemental materials to give readers additional information about their work. These materials have not been edited or formatted by Public Health Reports’s scientific editors and, thus, may not conform to the guidelines of the AMA Manual of Style, 11th Edition.

References

US Food and Drug Administration. FDA authorizes Pfizer-BioNTech COVID-19 vaccine for emergency use in children 5 through 11 years of age. News release. October 29, 2021. Accessed September 19, 2023. https://www.fda.gov/news-events/press-announcements/fda-authorizes-pfizer-biontech-covid-19-vaccine-emergency-use-children-5-through-11-years-age

US Food and Drug Administration. Coronavirus (COVID-19) update: FDA authorizes Moderna and Pfizer-BioNTech COVID-19 vaccines for children down to 6 months of age. News release. June 17, 2022. Accessed September 19, 2023. https://www.fda.gov/news-events/press-announcements/coronavirus-covid-19-update-fda-authorizes-moderna-and-pfizer-biontech-covid-19-vaccines-children

Szilagyi

Shah

Delgado

, et al Parents’ intentions and perceptions about COVID-19 vaccination for their children: results from a national survey. Pediatrics. 2021;148(4):e2021052335. doi:10.1542/peds.2021-052335

Rothman

Salovey

. Shaping perceptions to motivate healthy behavior: the role of message framing. Psychol Bull. 1997;121(1):3-19. doi:10.1037/0033-2909.121.1.3

O’Keefe

Nan

. The relative persuasiveness of gain- and loss-framed messages for promoting vaccination: a meta-analytic review. Health Commun. 2012;27(8):776-783. doi:10.1080/10410236.2011.640974

Kim

Sherman

Updegraff

. Fear of Ebola: the influence of collectivism on xenophobic threat responses. Psychol Sci. 2016;27(7):935-944. doi:10.1177/0956797616642596

Pokhrel

Bennett

Regmi

, et al Individualism–collectivism, social self-control and adolescent substance use and risky sexual behavior. Subst Use Misuse. 2018;53(7):1057-1067. doi:10.1080/10826084.2017.1392983

Dutta

. Communicating about culture and health: theorizing culture-centered and cultural sensitivity approaches. Commun Theory. 2017;17(3):304-328. doi:10.1111/j.1468–2885.2007.00297.x

Humphrey

Bliuc

. Western individualism and the psychological wellbeing of young people: a systematic review of their associations. Youth. 2022;2(1):1-11. doi:10.3390/youth2010001

10.

Vandello

Cohen

. Patterns of individualism and collectivism across the United States. J Pers Soc Psychol. 1999;77(2):279-292. doi:10.1037/0022-3514.77.2.279

11.

Trachtenberg

Haas

Kinney

Stanley

Krous

. Risk factor changes for sudden infant death syndrome after initiation of back-to-sleep campaign. Pediatrics. 2012;129(4):630-638. doi:10.1542/peds.2011-1419

12.

Rice

Atkin

, eds. Public Communication Campaigns. 4th ed. Sage; 2012.

13.

Maaravi

Levy

Gur

Confino

Segal

. “The tragedy of the commons”: how individualism and collectivism affected the spread of the COVID-19 pandemic. Front Public Health. 2021;9:627559. doi:10.3389/fpubh.2021.627559

14.

Jin

English

. Collectivism predicts mask use during COVID-19. Proc Natl Acad Sci U S A. 2021;118(23):e2021793118. doi:10.1073/pnas.2021793118

15.

Borah

. Message framing and COVID-19 vaccination intention: moderating roles of partisan media use and pre-attitudes about vaccination. Curr Psychol. Published online February 3, 2022. doi:10.1007/s12144-022-02851-3

16.

Courtney

Felig

Goldenberg

. Together we can slow the spread of COVID-19: the interactive effects of priming collectivism and mortality salience on virus-related health behaviour intentions. Br J Soc Psychol. 2022;61(1):410-431. doi:10.1111/bjso.12487

17.

Asfaw

Morankar

Abera

, et al Talking health: trusted health messengers and effective ways of delivering health messages for rural mothers in Southwest Ethiopia. Arch Public Health. 2019;77:8. doi:10.1186/s13690-019-0334-4

18.

Dada

Djiometio

McFadden

, et al Strategies that promote equity in COVID-19 vaccine uptake for Black communities: a review. J Urban Health. 2022;99(1):15-27. doi:10.1007/s11524-021-00594-3

19.

Haimowitz

Thompson

. Community members trusted by African American parents for vaccine advice. Hum Vaccin Immunother. 2019;15(7-8):1715-1722. doi:10.1080/21645515.2019.1581553

20.

Olson

Berry

Kumar

. Addressing parental vaccine hesitancy towards childhood vaccines in the United States: a systematic literature review of communication interventions and strategies. Vaccines (Basel). 2020;8(4):590. doi:10.3390/vaccines8040590

21.

Geiger

. Do people actually “listen to the experts”? A cautionary note on assuming expert credibility and persuasiveness on public health policy advocacy. Health Commun. 2022;37(6):677-684. doi:10.1080/10410236.2020.1862449

22.

Edwards

Hackell

; Committee on Infectious Diseases, Committee on Practice and Ambulatory Medicine. Countering vaccine hesitancy. Pediatrics. 2016;138(3):e20162146. doi:10.1542/peds.2016-2146

23.

Henrikson

Opel

Grothaus

, et al Physician communication training and parental vaccine hesitancy: a randomized trial. Pediatrics. 2015;136(1):70-79. doi:10.1542/peds.2014-3199

24.

Ruggiero

Wong

Sweeney

, et al Parents’ intentions to vaccinate their children against COVID-19. J Pediatr Health Care. 2021;35(5):509-517. doi:10.1016/j.pedhc.2021.04.005

25.

Palm

Bolsen

Kingsland

. The effect of frames on COVID-19 vaccine hesitancy. Front Polit Sci. 2021;3:661257. doi:10.3389/fpos.2021.661257

26.

US Food and Drug Administration. Pfizer-BioNTech COVID-19 vaccine EUA amendment review memorandum. April 9, 2021. Accessed March 15, 2022. https://www.fda.gov/media/148542/download

27.

Bernard

. Social Research Methods: Qualitative and Quantitative Approaches. 2nd ed. Sage; 2012.

28.

Zeger

Liang

. Longitudinal data analysis for discrete and continuous outcomes. Biometrics. 1986;42(1):121-130.

29.

Touloumis

. R package multgee: a generalized estimating equations solver for multinomial responses. J Stat Softw. 2015;64(8):1-14. doi:10.18637/jss.v064.i08

30.

Mervosh

. When vaccines aren’t an option: life for families with children under 12. The New York Times. September 2, 2021. Accessed April 22, 2022. https://www.nytimes.com/2021/09/02/us/parents-kids-vaccine.html%20

31.

Chen

. What’s holding up the COVID vaccines for children under 5? ProPublica. March 3, 2022. Accessed April 8, 2022. https://www.propublica.org/article/whats-holding-up-the-covid-vaccines-for-children-under-5

32.

Bever

Sellers

. Overwhelmed by chaos and uncertainty, families with kids under 5 are on a vaccine roller coaster. The Washington Post. February 21, 2022. Accessed April 8, 2022. https://www.washingtonpost.com/health/2022/02/21/parents-kids-under-5-vaccine

33.

Motta

Sylvester

Callaghan

Lunz-Trujillo

. Encouraging COVID-19 vaccine uptake through effective health communication. Front Polit Sci. 2021;3:630133. doi:10.3389/fpos.2021.630133

34.

Ashworth

Thunström

Cherry

Newbold

Finnoff

. Emphasize personal health benefits to boost COVID-19 vaccination rates. Proc Natl Acad Sci U S A. 2021;118(32):e2108225118. doi:10.1073/pnas.2108225118

35.

Calo

Gilkey

Shah

, et al Misinformation and other elements in HPV vaccine tweets: an experimental comparison. J Behav Med. 2021;44(3):310-319. doi:10.1007/s10865-021-00203

36.

Brewer

Gottlieb

Reiter

, et al Longitudinal predictors of human papillomavirus vaccine initiation among adolescent girls in a high-risk geographic area. Sex Transm Dis. 2011;38(3):197-204. doi:10.1097/OLQ.0b013e3181f12dbf

37.

Flood

Rousculp

Ryan

, et al Parents’ decision-making regarding vaccinating their children against influenza: a web-based survey. Clin Ther. 2010;32(8):1448-1467. doi:10.1016/j.clinthera.2010.06.020

38.

Gargano

Herbert

Painter

, et al Impact of a physician recommendation and parental immunization attitudes on receipt or intention to receive adolescent vaccines. Hum Vaccin Immunother. 2013;9(12):2627-2633. doi:10.4161/hv.25823

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.08 MB